DATA of the manuscript 'Modelling Desiccation-Induced Crack Porosity using a VG-based Shrinkage Model and Anisotropic Theory'

This data is exclusively for Catena's peer review process. The abstract of this manuscript is as follows:
Clayey soil undergoes shrinkage during desiccation, resulting in a transfer of porosity from within the soil aggregates into cracks. These cracks detrimentally affect the mechanical properties of the soil and alter its hydraulic behaviors by providing preferential pathways for fluid or contaminants. Accurate modelling of desiccation cracks is an essential prerequisite for comprehending crack mechanisms and predicting water movement. However, existing volume-based models often neglect soil subsidence and shrinkage anisotropy. To address this limitation, we developed a model that predicts crack porosity as a function of soil water content that incorporates shrinkage anisotropy. This model was founded on a unified set of governing equations describing the porosities of the matrix, subsidence, and cracking domains. Matrix porosity was estimated using a modified normalized VG-Peng shrinkage model, while shrinkage anisotropy was characterized by a geometric factor (rs) modeled with a logistic function. After validating both sub-models, the integrated crack porosity model that incorporates VG-Peng model and shrinkage anisotropy model was verified. The unified model was tested on both undisturbed and remolded soils using ring-knife samples of 200 cm³ and 500 cm³. The normalized VG-Peng equation exhibited an excellent fit to all experimental data (R² and NSE > 0.95). The evolution of shrinkage anisotropy displayed three distinct phases: initial vertical subsidence, followed by predominant vertical with slight horizontal shrinkage, and finally horizontal cracking. The crack porosity model accurately predicted experimental data across different layer thicknesses using anisotropic shrinkage inputs from both ring-knife types. By integrating shrinkage anisotropy into a volume-based framework, our model achieves higher predictive accuracy than previous approaches.